import pandas as pd
import numpy as np
import plotly
import seaborn as sns
import matplotlib.pyplot as plt
import plotly.express as px

Data Description

- VIN (1-10): The first 10 characters of the vehicle’s VIN (Vehicle Identification Number).
- County: The region where the vehicle’s owner is registered. Owners in Washington may have vehicles located in other states.
- City: The city where the owner lives.
- State: The state linked to the vehicle's registration.
- Postal Code: The 5-digit ZIP code of the owner’s address.
- Model Year: The year of the vehicle, decoded from the VIN.
- Make: The manufacturer of the vehicle, decoded from the VIN.
- Model: The vehicle’s model, decoded from the VIN.
- Electric Vehicle Type: Whether the vehicle is fully electric or a plug-in hybrid.
- CAFV Eligibility: Whether the vehicle qualifies as a Clean Alternative Fuel Vehicle (CAFV) based on Washington's 2019 House Bill 2042.
- Electric Range: How far the vehicle can go on electric power alone.
- Base MSRP: The lowest suggested retail price for the vehicle’s model.
- Legislative District: The Washington state legislative area where the owner lives.
- DOL Vehicle ID: A unique ID assigned by the Department of Licensing (DOL).
- Vehicle Location: The center of the vehicle's registered ZIP code.
- Electric Utility: The electric utility serving the owner’s address.
- 2020 Census Tract: The census tract code (state + county + tract) assigned by the U.S. Census Bureau.

df =  pd.read_csv("dataset.csv")
df.head()

	VIN (1-10)	County	City	State	Postal Code	Model Year	Make	Model	Electric Vehicle Type	Clean Alternative Fuel Vehicle (CAFV) Eligibility	Electric Range	Base MSRP	Legislative District	DOL Vehicle ID	Vehicle Location	Electric Utility	2020 Census Tract
0	JTMEB3FV6N	Monroe	Key West	FL	33040	2022	TOYOTA	RAV4 PRIME	Plug-in Hybrid Electric Vehicle (PHEV)	Clean Alternative Fuel Vehicle Eligible	42	0	NaN	198968248	POINT (-81.80023 24.5545)	NaN	12087972100
1	1G1RD6E45D	Clark	Laughlin	NV	89029	2013	CHEVROLET	VOLT	Plug-in Hybrid Electric Vehicle (PHEV)	Clean Alternative Fuel Vehicle Eligible	38	0	NaN	5204412	POINT (-114.57245 35.16815)	NaN	32003005702
2	JN1AZ0CP8B	Yakima	Yakima	WA	98901	2011	NISSAN	LEAF	Battery Electric Vehicle (BEV)	Clean Alternative Fuel Vehicle Eligible	73	0	15.0	218972519	POINT (-120.50721 46.60448)	PACIFICORP	53077001602
3	1G1FW6S08H	Skagit	Concrete	WA	98237	2017	CHEVROLET	BOLT EV	Battery Electric Vehicle (BEV)	Clean Alternative Fuel Vehicle Eligible	238	0	39.0	186750406	POINT (-121.7515 48.53892)	PUGET SOUND ENERGY INC	53057951101
4	3FA6P0SU1K	Snohomish	Everett	WA	98201	2019	FORD	FUSION	Plug-in Hybrid Electric Vehicle (PHEV)	Not eligible due to low battery range	26	0	38.0	2006714	POINT (-122.20596 47.97659)	PUGET SOUND ENERGY INC	53061041500

df.columns

Index(['VIN (1-10)', 'County', 'City', 'State', 'Postal Code', 'Model Year',
       'Make', 'Model', 'Electric Vehicle Type',
       'Clean Alternative Fuel Vehicle (CAFV) Eligibility', 'Electric Range',
       'Base MSRP', 'Legislative District', 'DOL Vehicle ID',
       'Vehicle Location', 'Electric Utility', '2020 Census Tract'],
      dtype='object')

Data Cleaning

Dropping the null values as they are less than 5% of missing values

df.isna().sum()/len(df)

VIN (1-10)                                           0.000000
County                                               0.000000
City                                                 0.000000
State                                                0.000000
Postal Code                                          0.000000
Model Year                                           0.000000
Make                                                 0.000000
Model                                                0.000178
Electric Vehicle Type                                0.000000
Clean Alternative Fuel Vehicle (CAFV) Eligibility    0.000000
Electric Range                                       0.000000
Base MSRP                                            0.000000
Legislative District                                 0.002539
DOL Vehicle ID                                       0.000000
Vehicle Location                                     0.000213
Electric Utility                                     0.003933
2020 Census Tract                                    0.000000
dtype: float64

df.dropna(inplace=True)

dropping the unwanted columns as they are contain values which are all unique and doesn't give any useful information for analysis

df.columns

Index(['VIN (1-10)', 'County', 'City', 'State', 'Postal Code', 'Model Year',
       'Make', 'Model', 'Electric Vehicle Type',
       'Clean Alternative Fuel Vehicle (CAFV) Eligibility', 'Electric Range',
       'Base MSRP', 'Legislative District', 'DOL Vehicle ID',
       'Vehicle Location', 'Electric Utility', '2020 Census Tract'],
      dtype='object')

df.shape

(112152, 17)

df.nunique()

VIN (1-10)                                             7522
County                                                   39
City                                                    435
State                                                     1
Postal Code                                             516
Model Year                                               20
Make                                                     34
Model                                                   114
Electric Vehicle Type                                     2
Clean Alternative Fuel Vehicle (CAFV) Eligibility         3
Electric Range                                          101
Base MSRP                                                30
Legislative District                                     49
DOL Vehicle ID                                       112152
Vehicle Location                                        516
Electric Utility                                         73
2020 Census Tract                                      1760
dtype: int64

import re

def xed(x):
    k,j = re.findall("\s.*",x)[0].strip(" (").strip(")").split(" ")
    return (float(k),float(j))

xed(df["Vehicle Location"][2])

(-120.50721, 46.60448)

df.columns

Index(['VIN (1-10)', 'County', 'City', 'State', 'Postal Code', 'Model Year',
       'Make', 'Model', 'Electric Vehicle Type',
       'Clean Alternative Fuel Vehicle (CAFV) Eligibility', 'Electric Range',
       'Base MSRP', 'Legislative District', 'DOL Vehicle ID',
       'Vehicle Location', 'Electric Utility', '2020 Census Tract'],
      dtype='object')

df["coordinates"] = df["Vehicle Location"].apply(lambda x: xed(x))

df.drop(columns=['DOL Vehicle ID','Vehicle Location'],inplace=True)

EDA

Univariate

df["County"].value_counts().head(10).plot(kind="bar",color="green")
plt.title("The top 10 counties by vechiles registered count")
plt.show()

> From the above we can see the top 10 counties where majority of people are using electronic vehicles as a transport means. The King county has the largest no.of vehicles registered.

df["Make"].value_counts().head(10).plot(kind="bar",color="red")
plt.title("The top 10 vehicle brands by vehicles in use count")
plt.show()

df["Make"].value_counts(ascending=True).head()

Make
TH!NK              3
BENTLEY            3
AZURE DYNAMICS     7
GENESIS           18
FISKER            19
Name: count, dtype: int64

> From the above we can observe that Tesla is dominating the electric vehicles industry with such a huge customer base, while some brands like THINK & BENTLY have such a low presence in the market.

df["Electric Vehicle Type"].value_counts().plot(kind="pie",color="yellow",autopct="%.2f",shadow=True,explode=[0.1,0])
plt.title("Distribution of Electric Vehicle types in-use")
plt.ylabel("")
plt.show()

> From the above we can observe that Battery based Electric Vehicles are more in use than Plug-in Hybrid based Electric Vehicles.

sns.histplot(df["Model Year"],kde=True,color="blue",bins=20)
plt.title("Distribution of vehicles manufactured year")
plt.show()

> From the above we can observe that majority of the people are using vehicles that were manufactured in the years: 2021 - 2022

sns.displot(df["Electric Range"],kde=True,color="purple")
plt.title("The distribution of the battery range of the vehicles")
plt.show()

df['Clean Alternative Fuel Vehicle (CAFV) Eligibility'].value_counts().plot(kind="pie",autopct="%.2f",explode=[0.02,0.02,0.01])
plt.title("Distribution of CAFV Eligibility")
plt.ylabel("")
plt.show()

df.groupby("Clean Alternative Fuel Vehicle (CAFV) Eligibility")["Electric Range"].agg(["min","max","mean","count"])

	min	max	mean	count
Clean Alternative Fuel Vehicle (CAFV) Eligibility
Clean Alternative Fuel Vehicle Eligible	30	337	163.822194	58395
Eligibility unknown as battery range has not been researched	0	0	0.000000	39097
Not eligible due to low battery range	6	29	19.363847	14660

> From the above, we can observe that around 48% of the cars have a battery range 0-30 which might be due to some vehicles not having the eligibility due to low battery percentage and others not having their battery range researched. The reason for that can be because of the vehicle being an old model which were manufactured before the electric vehicles era.

df["Model"].value_counts().head(10).plot(kind="barh",color="gold")
plt.title("The top 10 vehicle models by vehicles in-use count")
plt.show()

> From the above we can observe that MODEL 3, MODEL Y and LEAF are the models with highest demand in the market

df["Electric Utility"].value_counts().head(10).plot(kind="barh",color="navy")
plt.title("The top 10 power suppliers for the electric vehicle owners")
plt.show()

> From the above we can observe the top 10 electric power providers respective to the customers locations and usage.

temp = df[df["Base MSRP"]!=0] # removing the values 0 as the vehicles can't be sold for free. Considering that their values might be missing

sns.histplot(data=temp,x="Base MSRP",kde=True)
plt.title("The distribution of Base MSRP of the vehicles")
plt.show()

> From the above we can say that most of the vehicles have an Base MSRP of ranging between 0 to 100,000.

---

Bivariate

df.groupby("Make")["Model"].nunique().sort_values(ascending=False).head(10).plot(kind="barh",color="teal")
plt.title("No.of varaints available in each of the top 10 brands by count")
plt.show()

> From the above we can observe the top 10 brands which have a large range of variants available/being used in the market.

sns.histplot(data=df,x="Model Year",hue="Electric Vehicle Type",bins=20, multiple='stack')
plt.title("No.of vehicles manufactured in each year along with the type of electric vehicle")
plt.show()

> From the above plot we can observe that there was a cut in the manufacturing of vehicles in the year 2020 which might be due to the covid pandemic post which the production rose back up. The vehicles industry had a bloom in the 2019 and again in 2023.

df.groupby("Model")["Electric Range"].mean().sort_values(ascending=False).head(10).plot(kind="bar",color="olive")

<Axes: xlabel='Model'>

> From the above we can observe the top 10 vehicle models with a high average electric range.

temp = df[df["Electric Range"]!=0] # removing the values 0 as the battery based vehicles can't be having 0 Range. Assuming that their values might be missing.

sns.violinplot(data=temp,x="Electric Range",hue="Electric Vehicle Type")
plt.title("No.of vehicles manufactured in each year along with the type of electric vehicle")
plt.show()

> From the above we can see that the plug-in type vehicles mostly have a range of 0-50, while the battery type vehicles has a wide distribution from 170 to 300.

---

Create a Choropleth using plotly.express to display the number of EV vehicles based on location.

df["County"].unique()

array(['Yakima', 'Skagit', 'Snohomish', 'Island', 'Thurston', 'Grant',
       'King', 'Kitsap', 'Whitman', 'Spokane', 'Cowlitz', 'Pierce',
       'Kittitas', 'Grays Harbor', 'Clark', 'Chelan', 'Whatcom', 'Benton',
       'Walla Walla', 'Mason', 'San Juan', 'Lewis', 'Jefferson',
       'Clallam', 'Douglas', 'Klickitat', 'Skamania', 'Adams', 'Franklin',
       'Okanogan', 'Stevens', 'Asotin', 'Ferry', 'Pacific', 'Columbia',
       'Wahkiakum', 'Lincoln', 'Pend Oreille', 'Garfield'], dtype=object)

# df['County'] = df['County'] + " County"
temp = df.groupby("County")["VIN (1-10)"].count().reset_index()
temp.columns = ['County','EV Count']
county_fips = {
    'Adams': '53001', 'Asotin': '53003', 'Benton': '53005', 'Chelan': '53007',
    'Clallam': '53009', 'Clark': '53011', 'Columbia': '53013', 'Cowlitz': '53015',
    'Douglas': '53017', 'Ferry': '53019', 'Franklin': '53021', 'Garfield': '53023',
    'Grant': '53025', 'Grays Harbor': '53027', 'Island': '53029', 'Jefferson': '53031',
    'King': '53033', 'Kitsap': '53035', 'Kittitas': '53037', 'Klickitat': '53039',
    'Lewis': '53041', 'Lincoln': '53043', 'Mason': '53045', 'Okanogan': '53047',
    'Pacific': '53049', 'Pend Oreille': '53051', 'Pierce': '53053', 'San Juan': '53055',
    'Skagit': '53057', 'Skamania': '53059', 'Snohomish': '53061', 'Spokane': '53063',
    'Stevens': '53065', 'Thurston': '53067', 'Wahkiakum': '53069', 'Walla Walla': '53071',
    'Whatcom': '53073', 'Whitman': '53075', 'Yakima': '53077'
}
temp['FIPS'] = temp['County'].map(county_fips)

temp

	County	EV Count	FIPS
0	Adams	34	53001
1	Asotin	48	53003
2	Benton	1376	53005
3	Chelan	654	53007
4	Clallam	728	53009
5	Clark	6681	53011
6	Columbia	13	53013
7	Cowlitz	569	53015
8	Douglas	221	53017
9	Ferry	27	53019
10	Franklin	365	53021
11	Garfield	4	53023
12	Grant	335	53025
13	Grays Harbor	402	53027
14	Island	1298	53029
15	Jefferson	698	53031
16	King	58980	53033
17	Kitsap	3828	53035
18	Kittitas	392	53037
19	Klickitat	175	53039
20	Lewis	431	53041
21	Lincoln	30	53043
22	Mason	547	53045
23	Okanogan	149	53047
24	Pacific	145	53049
25	Pend Oreille	32	53051
26	Pierce	8525	53053
27	San Juan	717	53055
28	Skagit	1228	53057
29	Skamania	139	53059
30	Snohomish	12412	53061
31	Spokane	2785	53063
32	Stevens	91	53065
33	Thurston	4109	53067
34	Wahkiakum	39	53069
35	Walla Walla	312	53071
36	Whatcom	2839	53073
37	Whitman	177	53075
38	Yakima	617	53077

geojson_url = 'https://raw.githubusercontent.com/plotly/datasets/master/geojson-counties-fips.json'

fig = px.choropleth(
    temp, 
    geojson=geojson_url, 
    locations='FIPS',  # Use FIPS codes for matching
    color='EV Count', 
    scope='usa',  
    title='Number of Electronic vehicles in Washington State County-wise',
    labels={'EV Count': 'EV Count'}
)

fig.update_geos(fitbounds="locations", visible=False)  # Focus on WA state
fig.show()

Create a Racing Bar Plot to display the animation of EV Make and its count each year.

import bar_chart_race as bcr

temp = df.groupby(["Make","Model Year"])["VIN (1-10)"].count().reset_index(name='Count')
temp = temp[(temp["Model Year"]>=2012 )&(temp["Model Year"]<=2024)] # taking the data of past 10 years
temp = temp.sort_values(by="Model Year")

fig = px.bar(
    temp,
    x='Count',
    y='Make',
    color='Make',  # Optional: Color bars by Make
    animation_frame='Model Year',  # Use Year for the animation
    orientation='h',  # Horizontal bar plot
    title='EV Make Count Over the Years',
    labels={'Count': 'EV Count', 'Make': 'Vehicle Make'},
    range_x=[0, temp['Count'].max() + 100]  # Adjust range for better visualization
)

fig.update_layout(
    xaxis=dict(showgrid=True, gridcolor='LightGray', tickformat=',d'),
    yaxis_title='Make',
    xaxis_title='Number of Vehicles',
    showlegend=False,
    title_x=0.5,
    margin=dict(l=50, r=50, t=50, b=50),
    width=800,
    height=600
)
fig.show()

ev_counts = temp.pivot_table(index='Model Year', columns='Make', values='Count', fill_value=0)

# Pivot the DataFrame to the right format
ev_counts = temp.pivot_table(index='Model Year', columns='Make', values='Count', fill_value=0)

bar_chart_params = {
    'df': ev_counts,                             # DataFrame for the chart
    'filename': "ev_racing_bar.mp4",           # Output file name
    'orientation': "h",                         # Horizontal bar plot
    'sort': "desc",                             # Sort bars in descending order
    'n_bars': 10,                               # Display the top 10 bars
    'steps_per_period': 50,                     # Steps for smoothness
    'period_length': 2000,                      # Duration of each period in milliseconds
    'title': 'EV Make Count Over the Years' # Title settings
}

# Create the racing bar plot and save it as a video
bcr.bar_chart_race(**bar_chart_params)

# Confirmation message
print("Video saved as ev_racing_bar.mp4")

C:\Users\vijay\anaconda3\Lib\site-packages\bar_chart_race\_make_chart.py:889: FutureWarning:

Series.fillna with 'method' is deprecated and will raise in a future version. Use obj.ffill() or obj.bfill() instead.

C:\Users\vijay\anaconda3\Lib\site-packages\bar_chart_race\_make_chart.py:286: UserWarning:

set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.

C:\Users\vijay\anaconda3\Lib\site-packages\bar_chart_race\_make_chart.py:287: UserWarning:

set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.

C:\Users\vijay\anaconda3\Lib\site-packages\bar_chart_race\_make_chart.py:226: UserWarning:

Some of your columns never make an appearance in the animation. To reduce color repetition, set `filter_column_colors` to `True`

Video saved as ev_racing_bar.mp4